Download User Manual - Hydrometeorology and Remote Sensing Laboratory

NASA
University of Oklahoma (OU) HyDROS Lab (http://hydro.ou.edu)
CREST
Coupled Routing and Excess STorage
User Manual
Version 2.0
By Dr. Xianwu Xue and Dr. Yang Hong
University of Oklahoma, National Weather Center, Norman, OK USA
August 11, 2013
Cover: CREST—Coupled Routing and Excess Storage User Manual Version 2.0
Table of Contents
CREST ................................................................................................................ 1
TABLE OF CONTENTS ...................................................................................... I
1 INTRODUCTION ............................................................................................ 1
2 COMPILATIONS ............................................................................................. 3
2.1 Compiling on Linux .......................................................................................... 3
2.2 Compiling on Windows .................................................................................... 3
3 FRAMEWORK OF CREST V2.0 ..................................................................... 5
3.1 Programming Framework of CREST v2.0 ....................................................... 5
3.2 Organization of the files and folders ................................................................. 5
4 PROJECT FILE ............................................................................................... 9
4.1 Model Area........................................................................................................ 9
4.2 Model Run Time Information ......................................................................... 10
4.3 Model Run Style ............................................................................................. 11
4.4 Model Directory .............................................................................................. 11
4.5 OutPix Information ......................................................................................... 12
4.6 Outlet Information .......................................................................................... 13
4.7 Grid Outputs.................................................................................................... 13
4.8 Date Outputs ................................................................................................... 15
4.9 Number of Lakes............................................................................................. 15
5 INPUTS & OUTPUTS .....................................................................................17
5.1 Basics Folder ................................................................................................... 17
5.2 Params Folder ................................................................................................. 23
5.3 State Folder ..................................................................................................... 26
5.4 ICS Folder ....................................................................................................... 26
5.5 OBS Folder ..................................................................................................... 28
5.6 Calibs Folder ................................................................................................... 28
I
5.7 Rains Folder .................................................................................................... 31
5.8 PETs Folder ..................................................................................................... 31
5.9 Results Folder ................................................................................................. 31
6 RUN STYLES ..................................................................................................33
6.1 Simulation ....................................................................................................... 33
6.2 Automatic Calibration using SCE-UA ............................................................ 35
6.3 Simulation in real time mode .......................................................................... 38
6.4 Return Period mode ........................................................................................ 38
6.5 Forecast_KMDQPF mode .............................................................................. 40
7 IMPLEMENTATION OF THE CREST MODEL FOR OTHER BASINS ......43
8 CONTACT US .................................................................................................45
9 SELECTED CREST MODEL RELATED REFERENCES .............................45
10 APPENDIX A LOOK-UP TABLES ................................................................47
II
1 Introduction
The Coupled Routing and Excess STorage (CREST) distributed hydrological
model is a hybrid modeling strategy that was developed by the University of
Oklahoma (http://hydro.ou.edu) and NASA SERVIR Project Team (www.servir.net).
The CREST model was initially developed to provide real-time regional and global
hydrological prediction by simultaneously running over multi-basins with relatively
cost-effective computational efficiency (http://eos.ou.edu), however it is also very
applicable for small to medium size basins at high-resolutions. CREST simulates the
spatiotemporal variation of water and energy fluxes and storages on a regular grid
with the grid cell resolution being user-defined, thereby enabling multi-scale
applications (Figure 1-1). The scalability of CREST simulations is accomplished
through sub-grid scale representation of soil moisture storage capacity (using a
variable infiltration curve) and multi-scale runoff generation processes (using
multi-linear reservoirs). The representation of the primary water fluxes such as
infiltration and routing are physically related to the spatially variable land surface
characteristics (i.e., vegetation, soil type, and topography etc.). The runoff generation
component and routing scheme are coupled, thus providing realistic interactions
between lower atmospheric boundary layers, terrestrial surface, and subsurface water.
The above features enable CREST to be applicable at large global, regional, and
small catchment scales.
This user manual and the accompanying code provide a single basin example to
test the model. The CREST model can be forced by gridded potential
evapotranspiration and precipitation datasets (e.g. satellite-based precipitation
estimates, interpolated rain gauge observations, weather radar, and quantitative
precipitation forecasts from numerical weather prediction models). The flexible
simulation modes and embedded automated calibration algorithms make the CREST
a powerful yet cost-effective tool for distributed hydrological modeling and
implementation at a range of scale from globe, region, basin, to small catchment.
1
Figure 1-1 Core Components of the CREST model
(a) Vertical profile of a cell including rainfall-runoff generation, evapotranspiration, sub-grid cell
routing and feedbacks from routing; (b) variable infiltration curve of a cell; (c) plane view of cells and
flow directions; and (d) vertical profile along several cells including sub-grid cell routing,
downstream routing, and subsurface runoff redistribution from a cell to its downstream cells.
2
2 Compilations
CREST v2.0 is written in FORTRAN, and will run under most operating
systems. It has been successfully implemented on Pentium & PC based systems
(under Microsoft Windows and Linux).
It is not necessary to modify the source code of CREST in order to change
settings or switch to other basins. In CREST v1.6, the control file with a default
name of “control.txt” and basic grids dictate the necessary settings for running the
model. However, CREST v2.0 uses “ProjectName.Project” instead of the
“control.txt”. User can use project name as the Project file’s name, this makes the
Project file more readable.
2.1 Compiling on Linux
The Linux/Unix operating systems are case sensitive. So when you compile
CREST v2.0, you must pay attention to the name and extension of the default file.
2.1.1 Using “ifort” compiler
Compiling CREST is easy with FORTRAN compiler. The source code of
CREST model is contained in a single file for ease of use. As such, in order to
compile CREST using ifort all you need to do is to type a simple command line
"ifort crest_v2_0.for –o crest_v2_0.lx". This will compile the CREST FORTRAN
source code file into an executable named "crest_v2_0.lx". The Intel FORTRAN
compiler has many other command line arguments to enable additional optimizations
and other features. If you want a full list and description of how to use ifort, please
consult the Intel FORTRAN compiler user manual.
2.2 Compiling on Windows
2.2.1 Using “Compaq Visual FORTRAN” (CVF) compiler
Compiling CREST is also very easy using CVF on Windows platforms, you can
just open the “crest_v2_0.for” by CVF, and then compile it, and finally,
“crest_v2_0.exe” will be created. If you want a full list and description of how to
use CVF, please consult the Compaq Visual FORTRAN compiler user manual.
3
3 Framework of CREST v2.0
3.1 Programming Framework of CREST v2.0
Comparing to previous CREST v1.6c, the programming framework of CREST
v2.0 was redesigned to better suit for distributed hydrological modeling. As shown in
Figure 3-1, CREST v2.0 includes more spatially distributed input data (including a
prior parameters) and outputs more variables data.
Figure 3-1 Programming Framework of CREST v2.0
3.2 Organization of the files and folders
The previous version CREST v1.6c puts all data information into “Control.txt”
(Figure 3-2); this will make the control file too big when modelers want to add other
parameters or data into the CREST. Additionally, when implementation of the model
becomes complex, it will contain too many files under one file folder, causing
non-necessary confusion and inconvenience to users or modelers.
5
DDM.txt
FAM.txt
DEM.txt
Basics
Rain
PET
Control.txt
Discharge
Name.hyp.txt
Exports
Figure 3-2 Files’ and Folders’ Organization of CREST v1.6c
In CREST v2.0, control file was divided into “ProjectName.Project”,
“Parameters.txt”, “InitialConditions.txt” and “Calibrations.txt”. Each of four files is
put in standalone folders, including other related data and files. Thus the
“ProjectName.Project” file only contains the model’s input information and its
6
configuration. This will enables user to build and modify these files easily. Figure
3-3 shows all the folders defined in CREST v2.0 based on their functionality. More
detailed information of these folders will be discussed in following sections.
Basics
ICS
Params
OBS
States
ProjectName.
Project
Rain
PET
Results
Calibs
Figure 3-3 Files’ and Folders’ Organization of CREST v2.0
7
4 Project File
The file “ProjectName.Project” contains the information about Model Area, Run
Time Information, Configuration Directory, Run Style, Outputs Information for
Specified Pixels and Outlet, Outputs States and Outputs Date, and it also contains
file assignments and their formats (One line for each assignment or information).
The “ProjectName” is the name of the project, when run CREST v2.0 on
Linux/Unix operating system, the extension of project file should write as “Project”,
not “project” or others.
Note:
The statement in the project file can be listed in any order, but the keywords
should not be changed. The format of the statement is:
Keyword
=
Value
The statement appearing on the same line should be space- or tab-separated.
Comment lines must have a pound sign, #, in the first column.
Comment for the statement in the line must be placed after Value and be sure to
leave at least one space or tab between the Value and the comments.
Keyword is not case sensitive.
4.1 Model Area
Figure 4-1 Sample Model Area in “ProjectName.Project”
NCols:
Number of cell columns;
NRows:
Number of cell rows.
9
XLLCorner :
X coordinate of the origin (by lower left corner of the cell).
YLLCorner :
Y coordinate of the origin (by lower left corner of the cell).
CellSize:
Cell Size.
NoData_Value: The input values to be No Data in the input/output map file.
4.2 Model Run Time Information
Figure 4-2 Sample Model Run Time Information in “ProjectName.Project”
TimeMark:
The unit of time step. The possible units are “y” (year), “m”
(month), “d” (day), “h” (hour), “u” (minute), “s” (second).
TimeStep:
Time Step.
StartDate:
Start date of the simulation, its format is defined as
“yyyymmddhhuuss”, the length is up to the time step unit, for
example,
“yyyy”
when
time
step
unit
is
“y”;
“yyyymmddhhuuss” when time step unit is “s”
LoadState:
The mark for reading the state file. “yes” means user want to
run the model by state files as initial value, “no” means the
initial values are read determined by initial condition file.
WarmupDate:
Warm up date for the simulation, its format is defined the same
as “StartDate”.
EndDate:
End date for the simulation, , its format is defined the same as
“StartDate”.
SaveState:
The mark for saving the state file. “yes” means user want to
save the state files when finished running the model, “no”
means user does not want to save the state files..
10
4.3 Model Run Style
Figure 4-3 Sample Model Run Style in “ProjectName.Project”
The run style “simu” means simulation; other possible run styles are
“cali_SCEUA” (automatic calibration using SCE-UA method), “RealTime” (on line
mode), “repe” (return period) and “Forecast_KMDQPF” (Forecast using KMD QPF
Data) modes.
4.4 Model Directory
Figure 4-4 Sample Model Directory in “ProjectName.Project”
11
As shown in Figure 3-3, CREST v2.0 divides the input and output data into 9
groups, each groups has a standalone folders, such as “Basics”, “Params”, “States”,
“ICS”, “Rains”, “PET”, “Results”, “Calibs” and “OBS” (the name of the folder can
be user-specified, but the its keyword is fixed). Each folder contains some files
(detailed content will be introduced in the next chapter), the format of the folder
means all or most of the files in this folder will use this format. The file possible
formats of CREST v2.0 are "ASC", "TXT", "DBIF", "BIFFIT", "TRMMRT",
"TRMMV6", "NMQBIN", "ASBIMO" and "BIBIMO".
4.5 OutPix Information
Figure 4-5 Sample OutPix Information in “ProjectName.Project”
NOutPixs:
The number of output pixels
OutPixColRow: OutPixColRow is specified if the pixel is relative to the basic
grids or in latitude and longitude. A value of “yes” means the
location of the pixels is a column and row, a value of “no”
means the location is longitude and latitude.
OutPixNameX: The name of the Xth Pixels. The value of “X” is up to
NOutPixs (X = [1~ NOutPixs]).
OutPixLongX:
The longitude of the Xth Pixels when OutPixColRow is
assigned “no”.
OutPixLatiX:
The latitude of the Xth Pixels when OutPixColRow is assigned
“no”.
OutPixColX:
The Column of the Xth Pixels when OutPixColX is assigned
12
“yes”.
OutPixRowX:
The Row of the Xth Pixels when OutPixColX is assigned
“yes”.
4.6 Outlet Information
Figure 4-6 Sample Outlet Information in “ProjectName.Project”
HasOutlet:
Whether have outlet or not, a value of “yes” means research
area has an outlet, a value of “no” means have not.
OutletColRow: OutletColRow is specified if the outlet is relative to the basic
grids or in latitude and longitude. A value of “yes” means the
outlet of the pixels is a column and row, a value of “no” means
the outlet is longitude and latitude.
OutletName:
The name of the outlet Pixels.
OutletLong:
The longitude of the Outlet Pixels when OutletColRow is
assigned “no”.
OutletLati:
The latitude of the Outlet Pixels when OutletColRow is
assigned “no”.
OutletCol:
The Column of the Outlet Pixels when OutletColRow is
assigned “yes”.
OutletRow:
The Row of the Outlet Pixels when OutletColRow is assigned
“yes”.
4.7 Grid Outputs
Grid Outputs is the control of 2-D grid-based output, “yes” means output and
“no” means do not output. The run time of the model depends on the number of
13
outputs. A faster CREST model runtime can be achieved by reducing the number of
output variables. Output format is controlled by the ResultFormat in section 4.4. All
outputs are spatially interpolated to the proper resolution and clipped to either the
basic grids or the drainage area automatically.
Figure 4-7 Sample Grid Outputs in “ProjectName.Project”
GOVar_Rain:
The input precipitation; unit is mm/hour.
GOVar_PET:
The input PET; unit is mm/hour.
GOVar_EPOT: GoVar_PET*KE
GOVar_EAct:
The depth of simulated actual evapotranspiration; unit is
mm/hour.
GOVar_W:
The depth of water filling the pore space bucket "WM"
GOVar_SM:
Soil Moisture, a percentage of the WM and equals
GOVar_W/WM
GOVar_R:
The simulated discharge of each grid cell; unit is m³/s.
GOVar_ExcS:
The depth of surface excess rain; unit is mm/hour.
GOVar_ExcI:
The depth of interflow excess rain; unit is mm/hour.
GOVar_RS:
The depth of overland flow; unit is mm/hour.
GOVar_RI:
The depth of interflow flow; unit is mm/hour.
14
4.8 Date Outputs
Figure 4-8 Sample Output Dates in “ProjectName.Project”
Date Outputs is to be specified some dates what user is interested in.
NumOfOutputDates: The Number of Output Date
OutputDate_X:
The Xth Date to output, The value of “X” is up to
NumOfOutputDates (X = [1~ NumOfOutputDates]).
4.9 Number of Lakes
Figure 4-9 Sample Number Lake in “ProjectName.Project”
NumOfLakes is to be specified the number of Lakes in this research region.
NumOfLakes:
The Number of Lakes
15
5 Inputs & Outputs
CREST v2.0 can read all the Grid file formats (such as "ASBIMO", "BIBIMO",
"ASC", "TXT", "DBIF","BIFFIT", "TRMMRT", "TRMMV6" and "NMQBIN") and
can clip the file automatically when this file is not equal to the extent defined in
“ProjectName.Project”, so user does not need to extract the research area by
themselves as long as your prepared input files have broader space domain.
5.1 Basics Folder
This folder contains the basic file for the model, such as DEM file, FDR file
(Flow Direction), FAC file (Flow Accumulation) and so on.
5.1.1 DEM File
Required:
Always
Name:
DEM.*
Format:
All Formats
Purpose:
Contains a digital elevation model of the basin area, with heights in meters
Notes:
The coordinate system of CREST v2.0 can use both Geographic Coordinate
System and Projected Coordinate System (PCS)
5.1.2 FDR File
Required:
Always
17
Name:
FDR.*
Format:
All Formats
Purpose:
Contain a flow direction from each cell to its steepest downslope neighbor of the
basin area.
Notes:
The coordinate system of FDR File should be same as DEM File.
In the former version of CREST, direction coding only use the rule like
Figure 5-1 (a) generated by “CNT”, however, most popular software (like
ArcGIS) use the rule like Figure 5-1 (b), so CREST v2.0 uses both of the
two rules.
(a)
(b)
Figure 5-1 Flow Direction Coding
5.1.3 FAC File
Required:
Always
Name:
FAC.*
18
Format:
All Formats
Purpose:
Contains accumulation flow to each cell of the basin area.
Notes:
The coordinate system of FAC File should be same as DEM File.
5.1.4 Mask File
Required:
Optional
If omitted
If HasOutlet == “yes”, then CREST v2.0 will generate the mask filebased on
Outlet location.
If HasOutlet == “no”, then CREST v2.0 will generate the mask file based on
NoData_Value in DEM file.
Name:
Mask.*
Format:
All Formats
Purpose:
Contains a mask of the basin, indicating which cells in the other terrain are inside
the basin
Notes:
The coordinate system of Mask File should be same as DEM File.
5.1.5 GridArea File
Required:
Optional
19
If omitted, CREST v2.0 will generate the GridArea file based on the coordinate
system.
Name:
GridArea.*
Format:
All Formats
Purpose:
Contain the area of each cell in the basin
Notes:
The coordinate system of GridArea File should be same as DEM File.
5.1.6 AreaFact File
Required:
Optional
If omitted, CREST v2.0 will assign AreaFact.file a uniform value (1.00).
Name:
AreaFact.*
Format:
All Formats
Purpose:
Contains the area of each cell in the basin
Notes:
The coordinate system of AreaFact File should be same as DEM File.
5.1.7 Stream File
Required:
Optional, however if Stream file is omitted, Stream.def must be required.
If omitted, CREST v2.0 will read the threshold to determinate the stream from
Stream.def.
20
Name:
Stream.*
Format:
All Formats
Purpose:
Contain the information to show whether each cell is the stream.
Notes:
The coordinate system of Stream File should be same as DEM File.
5.1.8 Stream.def File
Required:
Optional, however if Stream.def is omitted, Stream file must be required.
Only use when Stream file omitted.
Name:
Stream.def
Format:
ASCII, only contain one value
Purpose:
Contain the threshold for determining the stream.
Notes:
Stream.def is used to compatible with the former version of CREST, instead of
Th parameter in the former version of CREST.
5.1.9 Slope File
Required:
Optional
If Slope omitted
If Slope.def exists, CREST v2.0 will calculate the Slope like the former
version.
21
If Slope.def does not exist, CREST v2.0 will calculate the Slope automatically.
Name:
Slope.*
Format:
All Formats
Purpose:
Identify the rate of maximum change in DEM file from each cell.
Notes:
The coordinate system of Slope File should be same as DEM File.
5.1.10 Slope.def File
Required:
Optional
Only use when Slope file omitted, CREST v2.0 will read this file. If Slope.def is
omitted yet, CREST v2.0 will calculate the Slope automatically.
Name:
Slope.def
Format:
ASCII, only contain one value
Purpose:
Contain the threshold for determining the stream.
Notes:
Slope.def is used to compatible with the former version of CREST, instead of
GM parameter in the former version of CREST.
5.1.11 Lake Mask File
Required:
Optional
If NumOfLakes==0 then
omitted
22
Name:
LakeMask.*
Format:
All Formats
Purpose:
Contains a mask of lakes in this region, indicating which cells is the lake
Notes:
The coordinate system of LakeMask File should be same as DEM File.
5.2 Params Folder
This folder contains all the configuration and values of parameters for the
model. There are total 18 parameters, classified into four types in this new version
(see Table 5-1).
CREST v2.0 puts the initial condition into ICS folder (will introduce later), TH
and GM are optional control parameters, and AreaFact can be calculated by ArcGIS
or other software, or by CREST v2.0 automatically. So, there are 12 essential
parameters in CREST v2.0. The look up tables, the range and the default value of
part/all of these parameters are listed in Appendices Table 10-1, Table 10-2 and Table
10-3.
23
Table 5-1 Classification in CREST v2.0 vs v1.6c
Module
Initial
condition
Physical
Parameters
Conceptual
Parameters
Adjustment
Parameters
Symbol (v2.0)
W0
SS0
SI0
Ksat
RainFact
WM
B
IM
KE
coeM
expM
Symbol (v1.6c)
iWU
iSO
iSI
pFc
Rain
pWm
B
pIM/100
pKE
coeM
expM
coeR
River
coeS
Under
KS
KI
LeakO
LeakI
(Omitted)/TH
TH
(Omitted)/GM
GM
AreaFact
AreaC
Description
Initial Value of Soil Moisture
Initial value of Overland Reservoir
Initial value of Interflow reservoir
the Soil saturate hydraulic conductivity
the multiplier on the precipitation field
The Mean Water Capacity
the exponent of the variable infiltration curve
Impervious area ratio
The factor to convert the PET to local actual
overland runoff velocity coefficient
overland flow speed exponent
multiplier used to convert overland flow
speed to channel flow speed
multiplier used to convert overland flow
speed to interflow flow speed
Overland reservoir Discharge Parameter
Interflow Reservoir Discharge Parameter
Threshold to determine which cells are river
cells
downstream cell is higher than the upstream
downstream cell is a nodata/outside region cell
multiplier that modifies the area of grid cells
5.2.1 Parameters.txt File
Required:
Always
Name:
Parameters.txt
Format:
ASCII
Purpose:
24
Contain the configuration for all parameters in CREST v2.0
Notes:
The statement in the “Parameters.txt” file can be listed in any order, but the
keywords should not be changed. The format of the statement is:
Keyword
=
Value
Such as:
SymbolType
=
Uniform/Distributed
If SymbolType
=
Uniform then
Symbol
=
Value
Else
Symbol.* file must be provided in the same folder.
End if
The statement appearing on the same line should be space- or
tab-separated.
Comment lines must have a pound sign, #, in the first column.
Comment for the statement in the line must be placed after Value and be
sure to leave at least one space or tab between the Value and the
Comments.
Keyword is not case sensitive.
5.2.2 “Sysmbol.*” File
Required:
Optional
Only needed when this parameter’s style is Distributed.
Name:
ParameterName.*
25
“ParameterName” possibly likes “Rain”
Format:
ASCII
Purpose:
Contain the configuration for all parameters in CREST v2.0
Notes:
The file coordinate system of Slope File should be same as DEM File.
5.3 State Folder
This folder contains the state files, such as “State_StartDate_SS0.*”,
“State_StartDate_SI0.*”and “State_StartDate_W0.*” when LoadState = “yes”.
When SaveState = “yes”, then “State_EndDate_SS0.*”, “State_EndDate_SI0.*”and
“State_EndDate_W0.*” will be created after running the model.
5.4 ICS Folder
This folder contains all the configuration and values of initial conditions for the
mode.
5.4.1 InitialConditions.txt File
Required:
Always
Name:
InitialConditions.txt
Format:
ASCII
Purpose:
Contain the configuration for initial conditions in CREST v2.0
Notes:
26
The statement in the “InitialConditions.txt” file can be listed in any order,
but the keywords should not be changed. The format of the statement is:
Keyword
=
Value
Such as:
SymbolType
=
Uniform/Distributed
If SymbolType
=
Uniform then
Symbol
=
Value
Else
Symbol.* file must be provided in the same folder.
End if
The Symbol includes “SS0”, “SI0” and “W0” (see Table 5-1).
The statement appearing on the same line should be space- or
tab-separated.
Comment lines must have a pound sign, #, in the first column.
Comment for the statement in the line must be placed after Value and be
sure to leave at least one space or tab between the Value and the
Comments.
Keyword is not case sensitive.
5.4.2 “Sysmbol.*” File
Required:
Optional
Only needed when this parameter’s style is Distributed.
Name:
ParameterName.*
“ParameterName” possibly likes “Rain”
Format:
ASCII
27
Purpose:
Contain the configuration for all parameters in CREST v2.0
Notes:
The file coordinate system of Slope File should be same as DEM File.
5.5 OBS Folder
This folder contains all the observed runoff data for the model calibration or
verification. The file’s name is name as “OutPixNameX_Obs.csv” and/or
“OutletName_Obs.csv” (“.csv” is the comma delimited file). “OutPixNameX” and
“OutletName” are the same as the project file.
5.6 Calibs Folder
This folder contains all the configuration and values of calibration for the
model.
5.6.1 Calibrations.txt File
Required:
Always
Name:
Calibrations.txt
Format:
ASCII
Purpose:
Contain the configuration for calibrations in CREST v2.0
Notes:
The statement in the “Calibrations.txt” file can be listed in any order, but
the keywords should not be changed. The format of the statement is:
Keyword
=
Value
28
The statement appearing on the same line should be space- or
tab-separated.
Comment lines must have a pound sign, #, in the first column.
Comment for the statement in the line must be placed after Value and be
sure to leave at least one space or tab between the Value and the
Comments.
Keyword is not case sensitive.
Figure 5-2 Sample of “Calibrations.txt” file
1. SCE-UA Parameters
iseed:
Initial random seed;
maxn:
Max no. of trials allowed before optimization is terminated
kstop:
Number of shuffling loops in which the criterion value must
chang by the given percentage before optimization is
terminated
29
pcento:
Percentage by which the criterion value must change in given
number of shuffling loops
ngs:
Number of complexes in the initial population
2. Configuration for calibration
NCalibStations: Number of Calibrated Stations
IsColRow:
Specified if the location of calibrated station is relative to the
basic grids or in latitude and longitude. A value of “yes” means
the outlet of the pixels is a column and row, a value of “no”
means the outlet is longitude and latitude.
3. Configuration for Each Station
Name_X:
The name of the Xth station
Value_X:
The region value of the Xth station
Long_X:
The longitude of the Xth station when IsColRow is assigned
“no”.
Lati_X:
The latitude of the Xth station when IsColRow is assigned
“no”.
Col_X:
The Column of the Outlet Pixels when IsColRow is assigned
“yes”.
Row_X:
The Row of the Outlet Pixels when IsColRow is assigned
“yes”.
Label_X:
The minimum, initial value and maximum of the Xth Label
parameter for calibration. The label name can see Table
5-1Error! Reference source not found.. The format of this
statement is:
Label_X
=
Min
Value
Max
X is the number of station to calibration.
Only required when user want to calibrate this parameter.
5.6.2 CalibMask.* File
Required:
Allways
Name:
30
CalibMask.*
Format:
All Formats
Purpose:
Contain the order number for calibration, when have more than ONE calibrated
station. The number depends on the regional number of each station.
Notes:
The coordinate system of CalibMask File should be same as DEM File.
5.7 Rains Folder
This folder contains the precipitation data, the format of file in the folder
depending on the RainFormat specified in the “ProjectName.Project” file. The
CREST v2.0 can clip the region defined in the “ProjectName.Project” file
automatically.
5.8 PETs Folder
This folder contains the potential evaporation data, the format of file in the folder
depending on the PETFormat specified in the “ProjectName.Project” file; and the
CREST v2.0 can clip the region defined in the “ProjectName.Project” file
automatically.
5.9 Results Folder
This folder contains the output files, the format of file in the folder depending
on the ResultFormat specified in the “ProjectName.Project” file.
31
6 Run Styles
Different run styles have different combinations of outputs; the following
section will introduce the outputs for each style.
6.1 Simulation
6.1.1 Running in Simulation Mode
To run the model in simulation mode the run style in the project file must be set
to "simu". Precipitation data and PET data are also needed for the simulation period.
By default CREST reads the “ProjectName.Project” located in the current working
directory. However, as of CREST v2.0 it is possible to specify the project file name
as a command line option to the CREST executable.
6.1.2 Simulation Mode Standard Outputs
Figure 6-1 Output results in screen for Simulation mode
33
The 1 line is the comment for requiring user to enter the name of the project file.
The 2 line is the user-typed line for the name and path of the project file.
The 3 line is the separation line.
The 4~6 line is the information of the CREST v2.0.
The 7 line is the separation line.
The 8 line is the start date and time for running the model.
The 9 line is the separation line.
The 10~21 line is the reading and writing data.
The 22 line is the separation line.
The 23 line is the information for running style to the current running model.
The 24~32 line is the reading and writing data.
The 33 line is the separation line.
The 34~37 line is the output results.
The 38 line is the separation line.
The 39 line is the end date and time for running the model.
The 40 line is the elapsed run time for running the model.
6.1.3 Simulation Mode File Outputs
Besides output the results to the screen, CREST v2.0 also outputs the results to a
log
file
in
the
same
folder
as
project
file,
“ProjectName_YYYY.MM.DD-HH.UU.SS_CREST.log”,
named
for
as
example:
“NZoia_2011.02.07-00.33.05_CREST.log”. The time in this file’s name depending
on the date and time the model is running. The log file will help the user record all
the things the user wants to see, even on the Linux operating system.
There are many files to output when the model is running in the simulation
mode:
Mask.*: When the mask file is omitted in the basics folder.
Slope.*: When the slope file is omitted in the basics folder.
Outlet_OutletName_Mask.*: The mask file for the specified outlet.
Outlet_OutletName_Results.csv: The results for the outlet location (Table 6-1).
(Regional
Mean
Value
based
Outlet_OutletName_Mask.*)
34
on
the
Table 6-1 Sample of Outlet_OutletName_Results.csv
DateTime
Rain
PET
EPot
EAct
W
SM
RS
RI
ExcS
ExcI
R
RObs
2003/1/1 0:00
0
0.215
0.043
0.023
38.451
0.534
1.114
1.452
0
0
0.605
0.605
2003/1/1 3:00
0
0.215
0.043
0.023
38.382
0.533
1.112
1.452
0
0
2.359
2.359
2003/1/1 6:00
0.009
0.215
0.043
0.024
38.331
0.533
1.107
1.452
0
0.002
3.375
3.375
2003/1/1 9:00
0
0.215
0.043
0.023
38.263
0.532
1.1
1.452
0
0
4.655
4.655
2003/1/1 12:00
0
0.215
0.043
0.023
38.194
0.531
1.089
1.452
0
0
6.218
6.218
2003/1/1 15:00
0.889
0.215
0.043
0.03
39.705
0.552
1.15
1.508
0.131
0.225
7.681
7.681
2003/1/1 18:00
0.018
0.215
0.043
0.025
39.668
0.551
1.137
1.509
0.001
0.004
8.794
8.794
2003/1/1 21:00
0
0.215
0.043
0.024
39.597
0.55
1.122
1.508
0
0
9.526
9.526
2003/1/2 0:00
0
0.215
0.043
0.024
39.526
0.549
1.108
1.508
0
0
10.018
10.018
Outlet_OutletName_ Results_Statistics.csv: The statistics for outlet location.
Table 6-2 Sample of Outlet_OutletName_Results_Statistics.csv
NSCE
1
Bias(%) 2.72E-03
CC
1
OutPix_OutPixName_X_Mask.*: The mask file for the specified output pixel.
OutPix_OutPixName_X_Results.csv: The results for the output pixel location.
(Regional
Mean
Value
based
on
the
Outlet_OutPixName_X_Mask.*)
OutPix_OutPixName_X_Results_Statistics.csv: The statistics for outlet
location (Table 6-2).
6.2 Automatic Calibration using SCE-UA
6.2.1 Running in Automatic Calibration Mode using SCE-UA
To run the model in automatic calibration mode using SCE-UA, the run style in
the project file must be set to "cali_SCEUA". Precipitation data, PET data and
observed discharge are all needed for the calibration period. A special feature such as
Reinitializing or Resuming Calibration has been included in CREST v2.0.
6.2.2 Automatic Calibration Mode Standard Output using SCE-UA
35
Figure 6-2 Output results in screen for cali_SCEUA mode
36
The 1 line is the comment for requiring user to enter the name of the project file.
The 2 line is the user-typed line for the name and path of the project file.
The 3 line is the separation line.
The 4~6 line is the information of the CREST v2.0.
The 7 line is the separation line.
The 8 line is the start date and time for running the model.
The 9 line is the separation line.
The 10~21 line is the reading and writing data.
The 22 line is the separation line.
The 23 line is the information for running style to the current running model.
The 24~32 line is the reading and writing data.
The 33 line is the separation line.
The 34~36 line is the parameters information for SCE-UA method.
The 37 line is the separation line.
The 38~51 line is the output results for each loop of SCE-UA method.
The 52~65 line is the output results simulated using the calibrated parameters.
The 66 line is the separation line.
The 67 line is the end date and time for running the model.
The 68 line is the elapsed run time for running the model.
6.2.3 Automatic Calibration Mode File Output using SCE-UA
Besides output the results to the screen, CREST v2.0 also outputs the results to a
log
file
in
the
same
folder
as
project
file,
“ProjectName_YYYY.MM.DD-HH.MM.SS_CREST.log”,
named
for
as
example:
“NZoia_2011.02.07-00.33.05_CREST.log”. The time in this file’s name depending
on the date and time the model ran. The log file will help the user record all the
things the user wants to see, even on the Linux operating system.
There are many files to output when the model is running in the “cali_SCEUA”
mode:
Mask.*: When the mask file is omitted in the basics folder.
Slope.*: When the slope file is omitted in the basics folder.
Outlet_OutletName_Mask.*: The mask file for the specified outlet.
Outlet_OutletName_Results.csv: The results for the outlet location (Table 6-1).
37
(Regional
Mean
Value
based
on
the
Outlet_OutletName_Mask.*)
Outlet_OutletName_ Results_Statistics.csv: The statistics for outlet location.
OutPix_OutPixName_X_Mask.*: The mask file for the specified output pixel.
OutPix_OutPixName_X_Results.csv: The results for the output pixel location.
(Regional
Mean
Value
based
on
the
Outlet_OutPixName_X_Mask.*)
OutPix_OutPixName_X_Results_Statistics.csv: The statistics for outlet
location.
SCEUAOut_YYYY.MM.DD-HH.UU.SS.dat: The results for SCE-UA method
Label.*: The calibrated parameters file. Labels shows in Table 5-1.
6.3 Simulation in real time mode
6.3.1 Running in Real Time Mode
To run the model in real time mode, the run style in the project file must be set
to "RealTime". Precipitation data and PET data are needed for the real time period.
6.3.2 Real Time Mode Standard Output
It is the same as Simulation mode, however, the end date and time depending on
the latest available precipitation and PET data or images.
6.3.3 Real Time Mode File Output
It is the same as Simulation, however, the end date and time is up to the
available precipitation and PET.
6.4 Return Period mode
6.4.1 Running in “repe” Mode
To run the model in “repe” (Return Period) mode, the run style in the project
file must be set to "repe". It needs run the model in “simu” mode with
GOVar_R=”yes” in advance, and then run the model in “repe” mode.
38
6.4.2 Real Time Mode Standard Output
Figure 6-3 Output results in screen for Return Period mode
The 1 line is the comment for requiring user to enter the name of the project file.
The 2 line is the user-typed line for the name and path of the project file.
The 3 line is the separation line.
The 4~6 line is the information of the CREST v2.0.
The 7 line is the separation line.
The 8 line is the start date and time for running the model.
The 9 line is the separation line.
The 10~19 line is the reading and writing data.
39
The 20 line is the separation line.
The 21 line is the information for running style to the current running model.
The 22~29 line is reading runoff data.
The 30 line is sorting for return period.
The 31 line is exporting file Num 1.
The 32 line is the separation line.
The 33 line is the end date and time for running the model.
The 34 line is the elapsed run time for running the model.
6.4.3 Return Period Mode File Output
Besides output the results to the screen, CREST v2.0 also outputs the results to a
log
file
in
the
same
folder
as
project
file,
“ProjectName_YYYY.MM.DD-HH.MM.SS_CREST.log”,
named
for
as
example:
“NZoia_2011.02.07-00.33.05_CREST.log”. The time in this file’s name depending
on the date and time the model ran. The log file will help the user record all the
things the user wants to see, even on the Linux operating system.
There is only one types of file to output when the model is running in the “repe”
mode:
Level.X.*: X is the level number, recording the return period values.
6.5 Forecast_KMDQPF mode
6.5.1 Running
in “Forecast_KMDQPF” Mode
To run the model in forecast KMDQPF mode, the run style in the project file
must be set to "Forecast_KMDQPF". Precipitation data and PET data are also
needed for the Forecast KMDQPF period mode.
6.5.2 Forecast KMDQPF Mode Standard Output
The same as the standard output of real time running mode, except of the
40
information for running style to the current running model.
6.5.3 Forecast KMDQPF Mode File Output
Besides output the results to the screen, CREST v2.0 also outputs the results to a
log
file
in
the
same
folder
as
project
file,
“ProjectName_YYYY.MM.DD-HH.MM.SS_CREST.log”,
for
named
as
example:
“NZoia_2011.02.07-00.33.05_CREST.log”. The time in this file’s name depending
on the date and time the model ran. The log file will help the user record all the
things the user wants to see, even on the Linux operating system.
41
7 Implementation of the CREST model for other basins
The CREST model automatically runs over the region defined by the project file.
Therefore, if you are operating with global basic grids it is possible to easily and
quickly model a basin in the world by just simply defining the outlet of a new basin.
In the event that the basic grids you are using do not cover the region which you
want to model or if you want to model a region with a finer resolution then it is
necessary to derive new project file. The steps to do this are described below.
Fully implementing the CREST model on any basin can be achieved in a three
step process:
1. A project file for the new modeling region needs to be created.
a. Copy an existing project file (even one of the provided example
project files) and modify the paths to point to the location of the
new paths of the folders.
b. Determine the latitudes and longitudes of a rectangle around the
region which you wish to model. These do not have to be precise
by any means and can be pulled from Google Earth or Google
Maps for an approximate region around the basin you wish to
model. However, the basin you wish the model on must be
entirely contained in the bounding rectangle you specified.
c. In order for CREST to work with the DEM, FDR and FAC files
produced by the CNT Tool or other software, they must be named
DEM.*, FDR.* and FAC.*.
2. The model can now be run for your new basin in any desired modes. To
get realistic results it is necessary to generate a prior parameters from
available land surface datasets or later calibrate the model using a
gauged station within your defined new basin. The automatic calibration
built into the CREST model is the easiest way to calibrate the model.
43
8 Contact us
Development and maintenance of the current official version of the OU-NASA
CREST model is conducted at the University of Oklahoma, Hydrometeorology and
Remote Sensing Laboratory (http://hydro.ou.edu) and Atmospheric Radar Research
Center (ARRC) located in the National Weather Center (http://nwc.ou.edu). For
information about the current release of the CREST model or to get the source code
for beta versions of releases under development, please send e-mail to Dr. Yang
Hong ([email protected]) and/or Dr. Xianwu Xue ([email protected]).
9 Selected CREST model Related References
Wang. J., Y. Hong, L. Li, J.J. Gourley, K. Yilmaz, S. I. Khan, F.S. Policelli, R.F.
Adler, S. Habib, D. Irwn, S.A. Limaye, T. Korme, and L. Okello, 2011: The
Coupled Routing and Excess STorage (CREST) distributed hydrological
model. Hydrol. Sciences Journal, 56, 84-98.
Xue X, Hong Y, Limaye AS, Gourley JJ, Huffman GJ, Khan SI, et al., 2013:
Statistical and hydrological evaluation of TRMM-based Multi-satellite
Precipitation Analysis over the Wangchu Basin of Bhutan: Are the latest
satellite precipitation products 3B42V7 ready for use in ungauged basins?
Journal of Hydrology, 499(0): 91-99.
Khan, S. I., Y. Hong, J. Wang, K.K. Yilmaz, J.J. Gourley, R.F. Adler,
G.R. Brakenridge, F. Policelli, S. Habib, and D. Irwin, 2011: Satellite
Remote Sensing and Hydrologic Modeling for Flood Inundation Mapping in
Lake Victoria Basin: Implications for Hydrologic Prediction in Ungauged
Basins, IEEE Transactions on Geosciences and Remote Sensing, 49(1), 85-95,
Jan. 2011, doi: 10.1109/TGRS.2010.2057513
Wu H, Adler RF, Hong Y, Tian Y, Policelli F., 2012: Evaluation of Global Flood
Detection Using Satellite-Based Rainfall and a Hydrologic Model. Journal of
Hydrometeorolog, 13(4): 1268-1284.
45
46
10 Appendix A Look-up Tables
Table 10-1 Look-up Table for UMD Vegetation Types
Value
0
1
2
3
4
5
6
7
8
9
10
11
12
13
UMD Vegetation Category
Water
Evergreen Needleleaf Forest
Evergreen Broadleaf Forest
Deciduous Needleleaf Forest
Deciduous Broadleaf Forest
Mixed Forest
Woodland
Wooded Grassland
Closed Shrubland
Open Shrubland
Grassland
Cropland
Bare Ground
Urban and Built
47
Rooting Depth
(m)
0.001
1
1.25
1
1.25
1.125
0.997
0.872
0.651
0.578
0.75
0.75
0.55
0.797
Table 10-2 Look-up Table for HWSD Soil Texture
Code
Texture
Abbr.
Fild
Capacity
θfc(m3/m3)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
No_Soil
Clay(heavy)
Silty Clay
Clay
Silty Clay Loam
Clay Loam
Silt
Silt Loam
Sandy Clay
Loam
Sandy Clay Loam
Sandy Loam
Loamy Sand
Sand
NS
CH
SIC
C
SICL
CL
SI
SIL
SC
L
SCL
SL
LS
S
0
0.36
0.36
0.36
0.34
0.34
0.32
0.3
0.31
0.26
0.33
0.23
0.14
0.12
Permanent
Wilting
Point θpw
(m3/m3)
0
0.21
0.21
0.21
0.19
0.21
0.165
0.15
0.23
0.12
0.175
0.1
0.06
0.04
Table 10-3 Range and Default value of Each Parameter
Params
RainFact
(l)
Ksat
(mm/d)
WM(mm)
B (l)
IM (l)
KE (l)
coeM
expM (l)
coeR (l)
coeS (l)
KS (l)
KI (l)
Min
Default
Max
0.5
1.0
1.2
0
500
1000
80
0.05
0
0.1
1
0.1
1
0.001
0
0
120
0.25
0.05
0.95
90
0.5
2
0.3
0.6
0.25
200
1.5
0.2
1.5
150
2
3
1
1
1
48
Hydraulic
conductivity
Ksat (cm/h)
0.00001
0.03
0.05
0.075
0.1
0.1
0.495
0.65
0.15
0.34
0.15
1.09
2.99
11.78